EUVIMEDEUVIMED entdecken
Crossref

FCOS: A Two-Stage Recoverable Model Pruning Framework for Automatic Modulation Recognition

Yao Lu, Tengfei Ma, Zeyu Wang, Zhuangzhi Chen, Dongwei Xu, Yun Lin, Qi Xuan, Guan Gui

IEEE Transactions on Cognitive Communications and Networking · 2026 · 12 · 10764-10778

DOI: 10.1109/tccn.2026.3706560 ↗

Kurz gefasst

Fachartikel in der Zeitschrift IEEE Transactions on Cognitive Communications and Networking, 2026. Verfasst von Yao Lu und 7 weiteren Autor:innen. Die Arbeit nennt 94 Literaturangaben. Gefördert von: National Natural Science Foundation of China, National Key Research and Development Program of China, Yangtze River Delta Science and Technology Innovation Community Joint Research Project.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

With the rapid development of wireless communications and the growing complexity of digital modulation schemes, traditional manual modulation recognition methods struggle to extract reliable signal features and meet real-time requirements in modern scenarios. Recently, deep learning based Automatic Modulation Recognition (AMR) approaches have greatly improved classification accuracy. However, their large model sizes and high computational demands hinder deployment on resource-constrained devices. Model pruning provides a general approach to reduce model complexity, but existing weight, channel, and layer pruning techniques each present a trade-off between compression rate, hardware acceleration, and accuracy preservation. To this end, in this paper, we introduce FCOS, a novel Fine-to-COarse two-Stage hybird pruning framework that combines channel-level pruning with layer-level collapse diagnosis to achieve extreme compression, high performance and efficient inference. In the first stage of FCOS, hierarchical clustering and parameter fusion are applied to channel weights to achieve channel-level pruning. Then a Layer Collapse Diagnosis (LaCD) module uses linear probing to identify layer collapse and removes the collapsed layers due to high channel compression ratio. Experiments on multiple AMR benchmarks demonstrate that FCOS outperforms existing channel and layer pruning methods. Specifically, FCOS achieves 95.51% FLOPs reduction and 95.31% parameter reduction while still maintaining performance close to the original ResNet56, with only a 0.46% drop in accuracy on Sig2019-12. Code is available at https://github.com/yaolu-zjut/FCOS

Förderung

  • National Natural Science Foundation of China
  • National Key Research and Development Program of China
  • Yangtze River Delta Science and Technology Innovation Community Joint Research Project

Zitieren

APA

Lu, Y., Ma, T., Wang, Z., Chen, Z., Xu, D., Lin, Y., Xuan, Q., & Gui, G. (2026). FCOS: A Two-Stage Recoverable Model Pruning Framework for Automatic Modulation Recognition. IEEE Transactions on Cognitive Communications and Networking, 12, 10764-10778. https://doi.org/10.1109/tccn.2026.3706560

Vancouver

Lu Y, Ma T, Wang Z, Chen Z, Xu D, Lin Y, et al. FCOS: A Two-Stage Recoverable Model Pruning Framework for Automatic Modulation Recognition. IEEE Transactions on Cognitive Communications and Networking. 2026;12:10764-10778. doi:10.1109/tccn.2026.3706560

BibTeX
@article{Lu2026,
  author = {Lu, Yao and Ma, Tengfei and Wang, Zeyu and Chen, Zhuangzhi and Xu, Dongwei and Lin, Yun and Xuan, Qi and Gui, Guan},
  title = {FCOS: A Two-Stage Recoverable Model Pruning Framework for Automatic Modulation Recognition},
  journal = {IEEE Transactions on Cognitive Communications and Networking},
  year = {2026},
  volume = {12},
  pages = {10764--10778},
  doi = {10.1109/tccn.2026.3706560}
}
Kostenlos starten →